1. Field of the Invention
This invention relates to a projection optical apparatus, and in particular to an apparatus for measuring the distortion or the magnification of a projection optical system.
2. Related Background Art
When measuring the distortion of an optical apparatus, particularly a projection optical system used in an exposure apparatus, a method using mark printing on a semiconductor wafer as shown, for example, in FIG. 1 of the accompanying drawings has heretofore been practised.
This method is one which utilizes an interferometer for measuring the coordinates position of a wafer stage to convert the amount of distortion of a projection optical system for exposure into a short dimension.
Referring to FIG. 1A, patterns P1 to P5 for distortion measurement are formed on a reticle R.
By the use of the reticle as described above, the patterns P1 to P5 are printed on a wafer W by whole surface exposure, as shown in FIG. 1B.
Subsequently, a blind is applied to the marks on the reticle R except for the central mark P3, and exposure is again effected on the wafer W as shown in FIG. 1C.
At this time, the coordinates value of the stage is set on the basis of the known design data, and as shown in FIG. 1D, the mark P3 is superposedly printed on the coordinates of the already printed marks PA to PE conforming to the respective design data thereof.
The marks superposedly formed in the manner described above will be congruent if it is assumed that there is no distortion in the projection optical system, but in fact, there is distortion in the projection optical system and therefore, deviation .DELTA. occurs between the marks. The amount of distortion of the projection optical system can be measured on the basis of this deviation .DELTA..
However, the method as described above requires the process of repeating the exposure and development of the wafer twice and thereafter effect measurement. This leads to the inconvenience that measurement takes a long time.
Also, errors in the stoppage accuracy of the stage, the pattern measurement accuracy, etc. directly affect the distortion measurement accuracy, and this leads to the inconvenience that the reproducibility of measurement cannot be said to be good.